linuxrustccryptographystartupthreat modelingkey managementsecure bootremote attestationroot of trustfirmware security
Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
ЗП не указанаразвернуть
На похожих ролях по рынку
$94к$137кмедиана$178к
44 в выборке
На СНГ рынке: $5.3к/мес ($3.6к - $7.9к)
90
Крутая вакансия
развернуть
Это четко описанная и важная роль в хорошо профинансированном AI hardware-стартапе с практической работой на уровне кремния, прошивок и Linux; главный компромисс - неопределенность и риски ранней стадии.
Ранняя стадия
Современный стек
AI-инфраструктура
Крупное финансирование
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Security Engineer (AI Hardware): Building security mechanisms for the control chain of AI devices and racks with an accent on secure boot, cryptographic identity, firmware updates, and Linux and embedded-system protection. Focus on designing threat models, root-of-trust architectures, attestation flows, and secure interconnects across bare metal, RTOS, Linux, and early hardware.
Location: London - JJ Mack or Bristol - The Crescent, United Kingdom
Company
hirify.global is a young AI hardware company building chips and systems designed to accelerate inference for large AI models.
What you will do
Define threat models, trust boundaries, and secure-by-design architecture across device, board, chassis, and rack controllers.
Design and build secure boot, roots of trust, verified and measured boot, signing, anti-rollback, and production debug controls.
Secure controller and host interconnects through authentication, integrity protection, and secure protocols.
Build cryptographic mechanisms for key hierarchies, provisioning, device identity, and remote attestation.
Secure firmware updates with signed images, staged rollout, rollback protection, and recovery.
Harden the Linux kernel and on-die firmware while collaborating with hardware and software teams on reviews, vulnerabilities, and fixes.
Requirements
Adversarial mindset and strong threat-modeling skills covering trust boundaries, assets, attacks, and residual risk.
Practical experience implementing secure boot, roots of trust, cryptography, and key management in software and silicon.
Experience designing secure protocols and authenticated channels between components.
Strong C and Rust skills, with a focus on memory safety and defensive low-level coding.
Experience building security mechanisms such as a secure-boot chain, key hierarchy, or attestation flow.
Experience working across bare metal, RTOS, Linux, hardware, and system-level security is relevant to the role.
Comfort working with incomplete information, evolving specifications, simulators, and early hardware.
Nice to have
Boot and runtime attestation or management-plane protocols such as DICE, IETF RATS, SPDM, MCTP, PLDM, or TDISP.
Experience with HSMs, TPMs, secure elements, Caliptra, secure debug, open firmware or boot projects, and image minimisation.
Supply-chain security, SBOM, SLSA, fuzzing, security testing, NIST 800-series, Common Criteria, OCP S.A.F.E., TCG, or DMTF standards.
Physical and side-channel security awareness, AI/LLM-assisted security tooling, or RISC-V, GPU, or ML-accelerator experience.
Culture & Benefits
Full-time role in a young company working across hardware and software on an evolving inference system.
Collaborative work with architecture, hardware, and software engineers across bare metal, RTOS, and Linux.
Private medical, dental and vision coverage, contributory pension, and life and critical illness insurance.
25 days of holiday plus bank holidays, meaningful equity, and an inclusive and diverse office environment.
Some responsibilities may require eligibility checks under UK, US, and other international export-control regulations.
Будьте осторожны
Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →